Thinning Algorithm

In this section, we elaborate on the operational principles of thinning algorithms. The primary objective of thinning is to transform connected regions into single-pixel-width connected curves that maintain the topological structure while centering along the region's medial axis. To achieve this, thinning algorithms perform iterative pixel-level operations including: 1) Removing isolated boundary pixels while preserving connectivity using 8-neighborhood analysis, 2) Applying morphological operations like hit-or-miss transforms to detect and process corner points, and 3) Implementing smoothing filters to eliminate jagged edges. These operations typically employ algorithms such as Zhang-Suen or Guo-Hall thinning methods, which use parallel processing of boundary pixels through multiple sub-iteration stages. The combination of these techniques ultimately converts connected regions into clean skeletal curves, effectively revealing the essential structural features of the original image regions while maintaining connectivity and central positioning.